How could a telegraph operator correct a mistake?

by namastexinxbed

Was there a backspace equivalent? Also, while I’m at it, was a second operator needed to ensure the message being sent was accurate and not sabotaged? Thanks in advance.

bovisrex

I can speak for Allied Communications Protocols up to the mid-90s. TTYs (teletypes) were a step up from telegraph keys as they had a keyboard, though the ones I trained on and used from '91 to '95 only had upper-case letters, and you had to hit a key to switch between letters and numbers/ symbols. The procedures used developed from those used on CW (constant wave -- Morse Code) circuits, and we had a bunch of ways to correct errors or request them, depending on how we were communicating and with whom. Because both Morse code and TTY circuits can be cumbersome, NATO, Allied groups, and various communication services shortened as much as they absolutely could.

For example, if you were on your ship (let's say "BVRX,"), and you were sending a line to someone on the USS REDDIT (RDDT,) asking the operator to repeat a message labeled "061040Z OCT 91" you'd say, using the Z signal "ZDK" asking for a retransmission). This is what you start typing on your teletype:

RDDT DE BVRX INT ZDK 061040Z K

As you're typing that line (the 'K' at the end is the same as saying "OVER" in a voice channel) your Chief comes up to you and says "Actually... the message we need is from 1030, not 1040." Since you haven't finished typing it yet, this is what you end up sending to the channel:

RDDT DE BVRX INT ZDK 061040Z EEEEEEEE 061030Z K

The eight Es indicate an error in the word or group of letters immediately before. If you had already sent the message, you could send the message again, with the prosign "C" after the identifying Date-Time Group (the 061040Z OCT 91 indicates a message written at 1040 Greenwich Mean Time, on October 6th, 1991) to indicate that the new message is a correction of the old one.

If you knew there was an error but didn't have the correction, you would still error-out the message:

RDDT DE BVRX INT ZDK 061040Z EEEEEEEE AR

The AR in that message means "Out," as in "I have ceased transmission and do not expect a reply.

One other set of prosigns that helped: WA and WB, which meant "Word After" and "Word Before." You could use these to request a correction ("RDDT DE BVRX INT WA SEND K") or to send one (BVRX DE RDDT WA SEND NEWS K") Note: the parenthesis and quotes are only used in my explanation. They, and indeed any punctuation not strictly necessary, were never included.

I'm not certain when these procedures were enacted, but they were used at least from the Korean Conflict on to the 1990s. The idea of typing a code to indicate a correction, though, is the same. As for the second part of your question: Depending on the sensitivity of the message, the classification of the circuit, and the type of message being sent, there were ways to check the message, such as transmitting a word count at the end of the message. This would be verified by the receiving station.

There are other prosigns and signals that can be used to tell the other operator, in as few letters as possible (this was in the days of 50 and 75 baud circuits) what you need to correct, or what you need correcting. The reference Allied Communications Pub 126 (ACP 126(C) UNCLASSIFIED) is the main reference for this. And though this is anecdotal, I can verify that the Radiomen on my first two ships were drilled on them so much that, even though I haven't operated a Navy comms circuit as my main job in 25 years (I moved on to the IT side of the house) I only had to look up one of the prosigns above. Everything else is ingrained in my memory, some of it in 5-position punch-code.

Sources:

ACP 126(C) MAY 1989

Q and Z signals.